MARCH 1996
RESEARCH PAPER FORTY-NINE
A STATISTICAL ANALYSIS OF
FOREIGN EXCHANGE RATE
BEHAVIOUR IN NIGERIA'S
AUCTION
GENEVESIO. OGIOGIO
AFRICAN ECONOMIC RESEARCH CONSORTIUM
CONSORTIUM POUR LA RECHERCHE ECONOMIQUE EN AFRIQUE
A statistical analysis of foreign
exchange rate behaviour in
Nigeria's auction
Other publications in the AERC Research Papers Series:
Structural Adjustment Programmes and the Coffee Sector in Uganda by Germina
Ssemogerere, Research Paper 1.
Real Interest Rates and the Mobilization of Private Savings in Africa by F.M. Mwega,
S.M. Ngola and N. Mwangi, Research Paper 2
Mobilizing Domestic Resources for Capital Formation in Ghana: the Role of Informal
Financial Markets by Ernest Aryeetey and Fritz Gockel, Research Paper 3.
The Informal Financial Sector and Macroeconomic Adjustment in Malawi by C. Chipeta
and M.L.C. Mkandawire, Research Paper 4.
The Effects of Non-Bank Financial Intermediaries on Demand for Money in Kenya by
S.M. Ndele, Research Paper 5.
Exchange Rate Policy and Macroeconomic Performance in Ghana by C.D. Jebuni, N.K.
Sowa and K.S. Tutu, Research Paper 6.
A Macroeconomic-Demographic Modelfor Ethiopia by Asmerom Kidane, Research Paper
7.
Macroeconomic Approach to External Debt: the Case of Nigeria by S. Ibi Ajayi,
Research Paper 8.
The Real Exchange Rate and Ghana's Agricultural Exports by K. Yerfi Fosu, Research
Paper 9.
The Relationship Between the Formal and Informal Sectors of the Financial Market in
Ghana by E. Aryeetey, Research Paper 10.
Financial System Regulation, Deregulation and Savings Mobilization in Nigeria by
A. Soyibo and F. Adekanye, Research Paper 11.
The Savings-Investment Process in Nigeria: an Empirical Study of the Supply Side by
A. Soyibo, Research Paper 12.
Growth and Foreign Debt: the Ethiopian Experience, 1964-86 by B. Degefe, Research
Paper 13.
Links Between the informal and Formal/Semi-Formal Financial Sectors in Malawi by
C. Chipeta and M.L.C. Mkandawire, Research Paper 14.
The Determinants ofFiscal Deficit and Fiscal Adjustment in Cote d'lvoire by O. Kouassy
and B. Bohoun, Research Paper 15.
Small and Medium-Scale Enterprise Development in Nigeria by D.E. Ekpenyong and
M.O. Nyong, Research Paper 16.
The Nigerian Banking System in the Context of Policies of Financial Regulation and
Deregulation by A. Soyibo and F. Adekanye, Research Paper 17.
Scope, Structure and Policy Implications of Informal Financial Markets in Tanzania by
M. Hyuha, O. Ndanshau and J.P. Kipokola, Research Paper 18.
European Economic Integration and the Franc Zone: the future of the CFA Franc after
1996. Part I: Historical Background and a New Evaluation of Monetary
Co-operation in the CFA Countries by Allechi M'bet and Madeleine Niankey,
Research Paper 19.
Revenue Productivity Implications of Tax Reform in Tanzania by Nehemiah E. Osoro,
Research Paper 20.
The Informal and Semi-formal Sectors in Ethiopia: a Study of the Iqqub, Iddir and Savings and Credit Co-operatives by Dejene Aredo, Research Paper 21.
Inflationary Trends and Control in Ghana by Nii K. Sowa and John K. Kwakye,
Research Paper 22.
Macroeconomic Constraints and Medium-Term Growth in Kenya: A Three-Gap
Analysis by F.M. Mwega, N. Nguguna and K. Olewe-Ochilo, Research Paper 23.
The Foreign Exchange Market and the Dutch Auction System in Ghana by Cletus K.
Dordunoo, Research Paper 24.
Exchange Rate Depreciation and the Structure of Sectoral Prices in Nigeria Under an
Alternative Pricing Regime, 1986-89 by Olu Ajakaiye and Ode Ojowu, Research
Paper 25.
Exchange Rate Depreciation, Budget Deficit and Inflation - The Nigerian Experience by
F. Egwaikhide, L. Chete and G. Falokun, Research Paper 26.
Trade, Payments Liberalization and Economic Performance in Ghana by C.D. Jebuni,
A.D. Oduro and K.A. Tutu, Research Paper 27.
Constraints to the Development and Diversification of Non-Traditional Exports in
Uganda, 1981-90 by G. Ssemogerere and L.A. Kasekende, Research Paper 28.
Indices of Effective Exchange Rates: A Comparative Study of Ethiopia, Kenya and the
Sudan by Asmerom Kidane, Research Paper 29.
Monetary Harmonization in Southern Africa by C. Chipeta and M.L. C. Mkandawire,
Research Paper 30.
Tanzania's Trade with PTA Countries: A Special Emphasis on Non-Traditional Products
by Flora Mndeme Musonda, Research Paper 31.
Macroeconomic Adjustment, Trade and Growth: Policy analysis using a Macroeconomic
model of Nigeria by C. Soludo, Research Paper 32.
Ghana: The Burden of Debt Service Payment Under Structural Adjustment by
Barfour Osei, Research Paper 33.
Short-Run Macroeconomic Effects of Bank Lending Rates in Nigeria, 1987-91:
A Computable General Equilibrium Analysis by D. Olu Ajakaiye, Research Paper
34.
Capital Flight and External Debt in Nigeria by S. Ibi Ajayi, Research Paper 35.
Institutional Reforms and the Management of Exchange Rate Policy in Nigeria by
Kassey Odubogun, Research Paper 36.
The Role of Exchange Rate and Monetary Policy in the Monetary Approach to the Balance of Payments: Evidence from Malawi by Exley B.D. Silumbu, Research Paper
37.
Tax Reforms in Tanzania: Motivations, Directions and Implications by Nehemiah
E. Osoro, Research Paper 38.
Money Supply Mechanisms in Nigeria, 1970-88 by Oluremi Ogun and Adeola Adenikinju,
Research Paper 39.
Profiles and Determinants of Nigeria's Balance of Payments: The Current Account
Component, 1950-88, by Joe U. Umo and Tayo Fakiyesi, Research Paper 40.
Empirical Studies of Nigeria's Foreign Exchange Parallel Market 1: Price Behaviour
and Rate Determination by Melvin D. Ayogu, Research Paper 41.
The Effects of Exchange Rate Policy on Cameroon's Agricultural Competitiveness by
Aloysius Ajab Amin , Research Paper 42
Policy Consistency and Inflation in Ghana by Nii Kwaku Sowa, Research Paper 43
Fiscal Operations in a Depressed Economy: Nigeria, 1960-90 by Akpan H. Ekpo and
John E. U. Ndebbio, Research Paper 44
Foreign Exchange Bureaus in the Economy of Ghana by Kofi A. Osei, Research Paper
45
The Balance of Payments as a Monetary Phenomenon: An Econometric Study of
Zimbabwe's Experience by Rogers Dhliwayo, Research Paper 46
Taxation of Financial Assets and Capital Market Development in Nigeria by Eno L.
Inanga and Chidozie Emenuga, Research Paper 47
The Transmission of Savings to Investment in Nigeria by Adedoyin Soyibo, Research
Paper 48
A statistical analysis of foreign
exchange rate behaviour in
Nigeria's auction
Genevesi O. Ogiogio
NISER
Ibadan, Nigeria
AERC Research Paper 49
African Economic Research Consortium, Nairobi
March, 1996
©African Economic Research Consortium, 1996
Published by The African Economic Research Consortium
P.O. Box 62882
Nairobi, Kenya
Printed by Majestic Printing Works Ltd.
P.O. Box 42466
Nairobi, Kenya
ISBN 9966-90Q-74-8
Contents
List of tables
Acknowledgement
I
Introduction
1
II The Management of the Exchange Rate in Nigeria
III Some Issues in the Analysis of Statistical Propoerties of Daily
Exchange Rates
IV The Research Methodology
V Behaviour of Exchange Rates in Nigeria's Foreign Exchange
Auction
VI Analysis of Effects
VII Summary and Conclusion
VII Policy Implications
2
13
28
33
35
References
36
9
11
List of tables
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
1986-1988 naira-dollar exchange rates
Components of Nigeria's external reserves (%)
Nigeria's foreign exchange market phases
Federal government budget deficits (1986-1990)
1991 budget performance (Nbn)
Summary of naira-dollar exchange rate indexes
Dispersal of biddings in the foreign exchange market
Summary indexes for Dutch auction (fortnightly biddings)
23 June 1988 - 8 December 1988
Summary indexes for inter-bank foreign exchange market
(daily biddings) 9 January 1989- 14 December 1990
Summary indexes for Dutch auction with daily biddings
14 December - March 1991
Summary indexes for Dutch auction with weekly biddings
March 1991 - October 1991
Summary of naira exchange rate indexes: All currencies
mean rate of change
Summary of naira exchange rate indexes: All currencies variance
rate of change
Summary of naira exchange rate indexes: All currencies rate of
depreciation/appreciation
Summary of naira exchange rate indexes: All currencies
exchange rate instability index
Summary of naira exchange rate indices: All currencies
coefficient of variation
Monthly naira-dollar exchange rate movements January December
Auction sysem and frequency of bidding effect on depreciation/
appreciation
Month-of-the year effect and the nature of distribution of the
rate of exchange rate changes 1986-1990
Day-of-the-week effect and the nature of distribution of the rate
of exchange rate changes 1986-1990
7
14
15
16
19
19
21
22
23
23
25
26
26
26
27
27
29
30
31
32
Acknowledgement
I acknowledge with thanks the financial suppport provided by the African Economic
Research Consortium (AERC) for this study and the very useful comments made by
AERC Resource Persons, including those on the Technical Committee, at the seminal
stage of this work. The views expressed in this research reflect neither those of the
AERC nor those of the National Centre for Economic Management and Administration.
All views are mine, so also are the remaining errors and omissions contained in this
report.
I. Introduction
j
For a developing country like Nigeria, the price of foreign exchange plays a highly
significant role in the ability of the economy to attain optimal productive capacity. Before
the dramatic change in exchange rate management policy in the wake of the economic
reform programme that began in July 1986, the supply of foreign exchange to the economy
was heavily subsidized through the overvaluation of the domestic currency (the naira).
In the years of abundant foreign exchange earnings, for example 1974-1980, the impact
of this subsidy was felt mainly on the consumption side. The manufacturing sector, also
benefitted, however, but agriculture suffered inadvertently. When the economic crisis
started in January 1981, efforts to preserve the subsidy became futile; the subsidy went
as economic rent to a handful of opportunists, and the expected benefits of a fixed rate at
less than market price were lost to the economy.
Given this situation in the face of rapidly declining foreign exchange earnings, there
was every reason for the government to seek to set the price for foreign exchange right
and provide incentives for agricultural and manufactured exports.
Since these efforts began, the monetary authorities have had to live, very
uncomfortably, with the puzzle surrounding the instability of the exchange rate. What is
considered a stable and realistic exchange rate has remained elusive. In the pursuit of
realism, the monetary authorities changed the auction system and frequency of bidding
in the official foreign exchange market on several occasions before the eventual collapse
of the auction system on 5 March 1992. Before this collapse, the following auction
systems were tried out:
Average rate pricing system with weekly bidding
Marginal rate pricing system with weekly bidding
Dutch auction system with fortnightly bidding
•
Inter-bank system with daily bidding
Dutch auction system with daily and subsequently weekly
bidding
9
J
9
The behaviour of the naira exchange rate under each system was markedly different
— volatile in some and fairly stable (but unrealistic) in others.
II. Management of the exchange rate in Nigeria
The concern with exchange rate management policy in Nigeria can be traced back to
1960 when the country became politically independent, even though the Central Bank of
Nigeria and the Federal Ministry of Finance had come into being two years earlier. There
are a variety of ways and reasons for classifying subsequent developments in Nigeria's
exchange rate policy. One classification would put such developments into two phases,
namely, a passive phase and an active phase. Yet another binary classification, which is
essentially equivalent to the foregoing, is the pre-structural adjustment and post structural
adjustment division in exchange rate policy developments. A classification that seems
to capture the historical sequence more closely, however, would divide exchange rate
management into five periods or stages.
Stage I: Fixed parity between the Nigerian pound and
the British pound
In the first period of exchange rate policy in Nigeria, which spanned the year 1960-1967
there was a one-to-one relationship between the Nigerian pound (N£) and the British
pound sterling (S£). This fixed parity lasted until the British pound was devalued in
1967. Because of the Nigerian civil war which was going on at the time, the monetary
authorities did not consider it expedient to devalue the Nigerian pound in sympathy with
the British pound.
Two major reasons accounted for this reluctance. First, a considerable share of the
country's resources was being diverted to finance the war. And second, there was the
apprehension that devaluation of the Nigerian pound would only raise domestic prices
of imports without any appreciable impact on exports, which were still largely primary
products. Rather than devalue the Nigerian pound, the monetary authorities decided to
quote the Nigerian currency in reference to the US dollar. In addition, they impose
severe restrictions on imports as well as strict administrative controls on foreign exchange.
Stage II: Fixed parity between the Nigerian pound and
the American dollar
The second period of Nigeria's exchange rate policy lasted from 1967 to 1974. During
A STATISTICAL ANALYSIS O F
FOREIGN EXCHANGE RATE BEHAVIOUR IN NIGERIA'S AUTCTION
3
this time a fixed parity was maintained with the American dollar. Nigeria abandonded
the US dollar and once again kept its currency at par with the British pound following
the international fincancial crisis of the early 1970s, which constrained President Nixon
to devalue the dollar.
It will be recalled that in 1971 the Bretton Woods (adjustable peg) system collapsed
and this gave rise to a new pattern of par value that was established by the Smithsonian
Agreement. This also eventally collapsed in March 1973, and led to the emergence of a
system of generalized floating of the major currencies of the international financial system.
Nigeria did not suddenly float its currency (now the naira) independently in line with the
generalized floating that emerged in 1973; rather, it retained the peg to the US dollar.
This policy did not last for long, however, because no sooner was the naira pegged to the
US dollar than the dollar was devalued.
During this phase of Nigeria's exchange rate policy it became apparent that there
were severe drawbacks in pegging the naira to a single currency. One clear case was that
the naira had to undergo a de-facto devaluation in sympathy with the dollar following
the devaluation of the latter. Initially the devaluation of the naira was thought capable of
ensuring stability of the local currency value of exports as well as protecting the local
industries from excssive external competition. It soon became clear to the monetary
authorities that pegging the naira to the weak US dollar, especially during the first quarter
of 1974, worsened Nigeria's inflationary situation. Against this background of mounting
inflationary pressures, the policy of pegging the naira to a single currency was abandoned
in 1974.
Stage III: Independent exchange rate policy
With the abandonment of the policy of pegging the naira to a single currency, between
April 1974 and late 1976, the Nigerian monetary authorities attempted an independent
exchange rate management policy that pegged the naira to either the US dollar or British
pound sterling, whichever currency was stronger in the foreign exchange market. The
main policy objective was to operate an independently managed exchanged system that
would influence real economic variables in the economy and bring down the rate of
inflation. Consequently, a policy of progressive appreciation of the naira was pursued
over the period; this was aided by the concurrent oil boom. Because of the huge earnings
from crude petroleum exports during this time Nigeria persistently ran appreciable external
surpluses in its balance of payments, which supported the appreciation of the naira.
The appreciation of the naira was, however, guided by:
changes in Nigeria's external reserves position;
changes in the balanace of payments; and
changes in the foreign markets of the currencies of
Nigeria's major trading partners, especially the US dollar and the British pound
sterling.
Given the circumstances that prevailed from 1974 to 1978, the policy of appreciating
9
9
9
4
RESEARCH PAPER 4 9
the naira exchange rate was considered appropriate since an over-valued currency was
necessary for implementing the import substitution industrialization programme that was
being pursued. This policy approach continued through 1976 when the economic fortunes
of Nigeria started to wane.
Stage IV: Pegging the naira to an import-weighted
basket o f currencies
Late in 1976, as a result of the changing fortunes of Nigeria's economy, a policy reversal
was effected in the management of the exchange rate. The naira was deliberately
depreciated, though this was not systematic. In the effort to realign the exchange rate,
the monetary authorities were convinced that a more appropriate way to ensure stability
and viability of the naira was to peg it to a basket of currencies. Hence, they adopted a
basket comprising the seven currencies of Nigeria's major trading partners. The currencies
were the US dollar, the British pound sterling, the German mark, the French franc, the
Dutch guilder, the Swiss franc, and the Japanese yen. This import-weighted basket
experiment was carried out between 1976 and 1985.
The experiment did lead to considerable stability in the naira exchange rate. However,
there was ample evidence following the economic crisis that started in January 1981,
and worsened progressively to 1985, that the exchange rate was unviable. Both in nominal
and real terms, the naira was grossly over-valued against the US dollar. Toward the end
of 1985, as the economic crisis deepened, the federal government was faced with two
alternatives in the management of the exchange rate. The first option was to continue
the policy of administratively fixing the exchange rate thus intensifying trade and exchange
controls. The second choice was to allow the rate to be determined by market forces.
The governement chose the second option and this led to the introduction and operation
of the second-tier foreign exchange market (SFEM), now foreign exchange market (FEM),
in September of 1986.
Stage V: Market determined exchange rate policy
This is the fifth stage in Nigeria's exchange rate management policy. It started off
effectively on 26 September 1986, when the country's exchange rate was first determined
officially through a public auction in the foreing exchange market. Since then, tremendous
progress has been made towards arrriving at a stable and viable exchange rate for the
naira. The current exchange rate policy is, indeed, the most significant policy reform in
Nigeria's economic adjustment programme. Every effort is therefore being made by the
monetary authorities to enusre that this phase in the country's exchange rate management
policy generates and establishes a realistic and sustainable rate for the naira. The evidence
so far, however has shown that there are still elements of distortion in the exchange rate
A STATISTICAL ANALYSIS O F
FOREIGN EXCHANGE RATE BEHAVIOUR IN NIGERIA'S AUTCTION
5
that make it difficult to ascertain what path of adjustment leads to stability and realism.
Operation of the foreign exchange market system
Under the market system, two exchange rates prevailed in two markets, described as
first-tier and second-tier foreign exchange markets. The first-tier rate applied to debt
servicing, payments to international imports for which letters of credit were opened before
the commencement of the second-tier market, and disbursements in respect of public
sector letters of credit.
In the second-tier market, little or no restriction was placed on the functioning of the
market. To ensure equity in the allocation of foreign exchange, banks were initially
classified into two categories, namely, big banks and small banks. Each successful big
bank could purchase a maximum of 10% and small banks a maximum of 7% of the total
foreign exchange allocation. Given this classification, it was clear that not all the
authorized dealers would be successful at each bidding session. Consequently, each
dealer tried to outbid the others by quoting high rates. Given the limited resources in the
market, the resultant high level of excess demand and aggressive posture of the banks,
the naira depreciated at the first and second bidding sessions when the average of
successful bids was used to determine the exchange rate.
After the second bidding, the marginal rate system was introduced as a mechanism
for setting the exchange rate in the belief that it would safeguard against the problem of
excessive depreciation. The marginal rate method enabled all successive bids to be
settled at the rate quoted by the last successful bank. Efforts at ensuring fairness and
viability in the market led to further amendments to the rules. Thus in order to increase
the number of successful bids, banks were reclassified into "big", "medium" and "new",
with maximum allocations of 5%, 2% and 1.5% respectively.
Although the naira firmed up at the end 1986 relative to its position at the beginning
of the second-tier market, the fluctuation from one bidding session to another was large.
The Central Bank of Nigeria actually had to intervene on two occassions in order to
moderate the amplitude of fluctuation in the exchange rate. The first intervention was
the sixth session when the initial foreign exchange supply was increased from $75 million
to $86 million in order to arrest the downturn in the rate. The second intervention, at the
tenth bidding session was meant to prevent an appreciation of the naira. These auctions
tended to stabilize the rate for some time.
During the first three months of 1987, however, the second-tier rate contined to
depreciate. It became clear that a change in the modality for determining the exchange
rate was necesssary so as to preserve the gains already made. Consequently, the Dutch
auction system was introduced in April 1987. Under this system, bids were settled at bid
rates in order to check unrealistic bids that had no bearing with actual demand. Under
the Dutch auction there were as many rates as there were successful bidders. However,
one worrisome development was that the rates diverged widely from one another, implying
the existence of distortions in the allocation of resources. Furthermore, after the initial
6
RESEARCH PAPER 4 9
show of strength, the naira weakened again and by the end of 1987, when it stood at
US$1.00 = N4.1413, it had depreciated by 19.9%.
Meanwhile, the inter-bank foreign exchange market had become important and added
another source of instability to the naira exchange rate. Initially, the inter-bank rate was
linked to the second-tier rate but later the link was broken. The banks were then permitted
to buy and sell foreign exchange at the rate determined in that market subject only to the
condition of a 1% spread between their buying and selling rates.
The inter-bank rate also diverged widely from the first tier market rate, which implied
that resources were not being rationally allocated. This caused a lot of concern to the
monetary authorities. Data on inter-bank rates showed that in July 1987, the rate stood at
US$1 =N4.3550 compared with US$1 =N3.8081 in the foreign exchange market (Table
1). Although the differential narrowed during 1987, it widened progressively to 38% by
April 1988 and by December 1988 stood at 54.8%. By the end of 1988, therefore, it was
clear that the multiplicity of rates under the Dutch auction system as well as the large
differential between the foreign exchange market and autonomous rates, was exerting
undue pressure on the exchange rate and intensifying the problem of resource
misallocation. Indeed, by the end of 1988, the naira had depreciated on the FEM from
US$1 = N1.5691 at the commencement of the second-tier market to N5.3530, reflecting
a 70.7% depreciation (Table 1). The autonomous rate itself had depreciated from US$1
= N3.7375 in July to US$1 = N8.2856 at the end of 1988 (54.9%). The rate at the end of
1988 indicated a premium of N2.9326 or 54.8% on the FEM rate.
In order to redress the situation and establish a single naira exchange rate, the
autonomous market and the official foreign exchange market were merged with effect
from January 1989 to form the inter-bank foreign exchange market (IFEM). No distinction
was made in the inter-bank market between funds obtained from the Central Bank and
those from autonomous sources. The exchange rate was determined by the Central Bank
using one or a combination of the following guiding principles:
(a) Weighted average of all quotations submitted by banks (individual
bank quotation weighted by amount demanded).
(b) Simple average of all quotations submitted by all banks.
(c) Highest and lowest bank quotes, provided the latter does not depreciate by
more than 2% when compared with the rate that emerges in (b) above.
(d) Intelligence report on the exchange rate movements during the previous day in
both the inter-bank foreign exchange market and some world financial centres.
Under the inter-bank market, which lasted until 14 December 1990, the monetary
authorities succeeded in establishing a single and fairly stable exchange rate for the naira
in the official market. This is one measure of success in the operation of the market.
However, the weakness of the naira continued to be an issue of concern to the monetary
authorities. For example, the exchange rate stood atN6.7178 = US$1 during the month
of January 1989 but had depreciated to N7.5871 by March. The rate strengthened
progressively from N7.5808 = US$1 in April to N7.1388 = US$1 in July 1989, after a
US$1 in August compared with N7.0389 = US$1 in January 1989.
In order to ensure adequate funding of the foreign exchange market and thereby ease
Table 1: 1986 -1990 naira-dollar exchange rates
Month
SEEM
Parallel
market
January
February
March
April
May
June
July
August
September
4.6406
October
4.1203
November
3.5311
December
3.1828
3.8687
Average
Depreciation/
Appreciation -45.8024
Source: Central Bank of Nigeria.
Firsttier rmarket
1
1
1
1.02
1.03
1.12
1.27
1.33
1.44
1.83
2.29
2.59
1.41
61.3896
SEEM
3.647
3.7014
3.9218
3.9054
4.1617
4.0506
3.8081
4.0809
4.2767
4.2767
4.289
4.1664
4.018025
12.4664
Parallel
market
Firsttier market
2.61
2.92
3.25
3.44
3.51
3.64
3.8081
3.311157
31.46188
Autonmous
SEEM
4.355
4.4579
4.4934
4.5694
4.553
4.5291
4.49267
4.1748
4.2611
4.3169
4.2023
4.1103
4.1913
4.6087
4.583
4.7167
4.7748
5.1479
5.343
4.5359
3.844031
21.86412
1988
Parallel Autonomous
market
market
4.449
4.86
4.5488
4.5
4.7214
4.3
5.3147
4.3
5.4
6.1539
6.3597
5.9
6.2765
6.35
6.6427
6.6
6.5608
6.2
6.0317
7.33
7.5
6.4761
8.2856
8.35
5.965833 5.985075
41.79641
46.30443
8
RESEARCH PAPER 4 9
series of tight monetary policy actions had been taken. The rate averaged n7.2593 = the
pressure on the naira, various balance of payments support loans were arranged by the
government. These included a World Bank trade policy and export development loan
($452 million). An Organization of Economic Cooperation Fund loan from Japan ($100
million) and a British government loan of $100 million. Another method by which the
government has been trying to improve the allocative efficiency of the foreign exchange
market has been through the establishment of bureaux de change. The aims of the policy
have been to allow small dealers in foreign exchange more freedom of action and to
enlarge the size of the officially recognized market and so improve macroeconomic
management.
Despite these developments, the exchange rate is yet to attain a stable level. In the
continued quest for stability and realism to the exchange rate, the monetary authorities
on 14 December 1990 abandoned the bidding system and method for determining the
exchange rate under the inter-bank market and re-introduced the Dutch auction system
with daily bidding sessions. By March 1991 bidding on a daily basis had been replaced
by bidding twice weekly. The consequence of the change to the Dutch auction system
has been another sudden depreciation of the naira. As at December 1990, the naira was
exchanging for the dollar at the rate of N8.71: US$ 1.00 whereas by February 1991 it had
depreciated to N9.40: US$1.00, reflecting a depreciation of 7.45%.
III. Some issues in the analysis of statistical
properties of daily exchange rates
The analysis of the statistical properties of daily exchange rates is important for
understanding the possible causes of instability or otherwise in a country's marketdetermined exchange rate management system. In a market system, it is observable that
the distribution of daily spot exchange rate changes differs between weekdays and
•weekends. In Nigeria, a close look at the movement of exchange rates in the parallel
market, for instance, shows that there is a tendency for the naira exchange rates to
appreciate on Fridays but depreciate on Mondays and Tuesdays.
It has also been observed that the impact of daily and fortnightly biddings in the
formal foreign exchange markets differs significiantly in terms of the ruling exchange
rates they generate and the rate of change of such exchange rates. Such difference is
emphasized, more or less, depending on whether the biddings are done under a Dutch
auction system or margianal or average rate pricing systems.
Studies on exchange rate behaviour have in the past tested for types of distribution
and "day-of-the-week" effects among others. The "day-of-the-week" effect is usually to
ascertain whether the distribution of daily spot exchange rates is different across days of
the week. In the literature, there is consensus that the distribution of the daily changes in
rates is unimodal and has fatter tails than the normal distribution. See, for instance Burt,
Kaen and Booth (1977), Westerfield (1977), Rogalski and Vinso (1978), and Hsieh (1988)'.
It is this second feature of the distribution that has attracted the most attention. Two
competing explanations have been advanced to explain the leptokurtosis. One view
suggests that the data were independently drawn from a fat tail distribution that remains
fixed over time, while the other proposes that the data come from distribution that vary
over time.
Few attempts have been made to determine which of the two explanations better
characterizes the data. Friedman and Yandersteel (1982) considered three hypotheses,
namely:
(1) The data are independent and identically distributed (IID), drawn from a table
symmetric paretian distribution.
(2) The data are IID, drawn from a mixture of two normal distributions.
(3) The data are independently drawn from a normal distribution whose
mean and variance change over time.
They found evidence in favour of the third hypothesis. Calderon-Rossel and benHorim (1982) tested directly for the IID property. They found that they can reject IID for
most currencies. They attribtued the rejection to shifts in the means of 9 out of 13
10
RESEARCH PAPER 4 9
currencies, and to changes in the skewedness in the remaining cases. Hsieh (1988)
found that exchange rate changes are not independent and identically distributed. Each
day of the week may have a different distribution, but this is not sufficient to explain the
rejection of IID; there is little serial correlation in the data and the means and variances
change over time.
Our concern in this study is not simply to examine the statistical properties of the
exchange rates as they relate to the three hypotheses put forward by Friedman and
Vandersteel (1982). Rather, the main thrust of our work is to examine these properties as
they contribute to the relative frequency of the depreciation, appreciation or stability of
the exchange rates.
IV. The research methodology
Data requirement
The data for the study, which were obtained from the Central Bank of Nigeria (CBN),
consist of daily weekly and fortnightly closing bid rates for the following currencies:
• US dollar (US$)
• British pound sterling (S£)
® Japanese yen (yen)
German mark (DM)
Dutch guilder (DG)
French franc (FF)
Swiss franc (SWF)
The following auction systems with their corresponding frequencies of bidding sessions
form the basis of the analysis:
Average rate system (with a second tier foreign exchange market) weekly bids
Marginal rate system (with a second tier foreign exchange market) - weekly
bids
Dutch auction system (with autonomous market) - fortnightly bids
Inter-bank foreign exchange system - daily bids, but with ruling rates determined
by the Central Bank
Dutch auction system - daily and weekly bids
9
9
9
9
9
9
9
9
Determination of the rate of change in exchange rates
Following Hsieh (1988), our study calculated the rate of change for each currency by
taking the logarithmic difference between the close of two successive trading days or
weeks. Thus, for exchange rate y, the rate of exchange was defined as follows:
Exchange rate change (ERC) - log (Yt/Yt-1) x 100
Based on the values obtained for ERC, the average or mean rate of change and the variance
of the rate of change were obtained.
12
RESEARCH PAPER 4 9
Determination of the coefficient of variation
For ease of comparison of relative dispersion or variability across currecies, the study
calculated the coefficient of variation (COV) for each of the currencies over the various
phases of the auction system. The COV was defined as follows:
Coefficient of variation (COV) = Variance of v
Mean of y
v
Measurement of relative stability of exchange rates
In order to verify further the issue of relative stability of the naira-dollar exchange rates
under the various auction systems, we look at the behaviour of an index of instability
that was derived as follows:
Exchange rate instability index (ERI) = VVar (y) x /
Where,
Var (y) = The variance of y
f = The mean rate of change of y
f=
i=i
f (ERC)
j
i
One advantage of this index is that it combines the variance of y and its rate of change
into a single statistic that reports the extent to which that variable is stable. This avoids
the problem of assessing the relative stability of the exchange rate data series using the
coefficient of variation (adjusted variance), which does not capture the size of the rate of
exchange of that series, or using the variance of the rate of change, which misses the
variability in the series.
V. Behaviour of exchange rates in Nigeria's
foreign exchange auction
Demand-supply indexes and bids for foreign exchange
in the foreign exchange auction
In Nigeria demand and supply indicators as well as bids for foreign exchange in the
foreign exchange auction are usually expressed in US dollars the principal currency in
this context. Although the demand and sypply situations are expressed in dollars, these
indicators are actually an expression of the economy's requirement for, and capacity to
generate, foreign exchange. Thus the demand for and supply of currencies such as pound
sterling, deutsche mark, Dutch guilder, French franc, Swiss franc and Japanese yen are
all measured in US dollar equivalents.
There may be several reasons for the use of the dollar as the lead currency. A major
one is that the bulk of foreign exchange supply to the foreign exchange market (FEM) is
in dollars (through crude oil export earnings). Moreover, since the 1980s, US dollar
assets have constituted the largest componet of Nigeria's external reserves (Table 2).
Hence the operations of the auction are conducted in dollars. However, having arrived
at the ruling rate for the dollar, the monetary authorities have to determine the exchange
rates for the other currencies relative to the dollar in major financial centres in Europe
and the United States.
What this implies is that, apart from the US dollar, the naira exchange rate for a
currency, say the Japanese yen, is not directly determined in the foreign exchange auction.
The movement of exchange rates in the auction is therefore expected to differ across
currencies unless exchange rate parities are fixed by countries having convertible
currencies. What this suggests in the context of this study is that the analysis of the/
properties of the exchange rates will inevitably focus more attention on the behaviour of
the naira-dollar exchange rate whenever there is need to make a judgement.
Phases in the foreign exchange auction system and
macroeconomic policy characteristics
Since the commencement of the public auction for the determination of the exchange
rate for the naira vis-a-vis the dollar (and some other major currencies), the value of the
Table 2: Components of Nigeria's external reserves (%)
Component 1970
1971 1972
Old IMF
Old
Sterling
Assets
Assets
Assets
French
franc
Canadian
dollar
Japanese
yen
Swiss
franc
Belgian
franc
Dutch
gilder
Others
7.9
7.9
15.4
75.4 73.9
9.2 10.3
1973 1974
1975 1976
1977 1978
1989 1990 1981 1982 1983
9.6
13.4
10.7
9.5
1.1
1.1
1.2
1.2
1.4
1.7
1.3
2.2
27.1
4.6
54
23
59.1
20.4
0.3
57.6
38.8
1.3
33.1
41.8
12.5
14.7
20.9
26.6
16.5
15.8
30.5
14
14.6
14.8
0.1
3.9
6
4
5
7.7
6.8
4.5
15
12.7
2.1
2
1.4
7.3
9.2
4.5
5.8
1.5
1.9
3.5
3.9
1.4
1.4
0.6
Source: Central Bank of Nigeria.
•
7.8
2.6
2.6
3.6
1985
1986
1987
19881989
1.8
2.7
2.4
10.2
6.5
0.5
1.1
0.5
0.5
0.4
15.1 18.8
19.7 39
13.2
6.3
17.3
33.9
5.4
16.6
51.2
6.9
7.1
58.2
6.2
4.9
76.2
5.4
9.1
69.6
5
13.3
1.1
0.2
9.3
3.5
8
5
2.1
1
0.9
0.6
0.6
0.3
0.2
0.2
9.1
9.6
0.5
0.3
6.4
6.8
9.4
12.8
23.4
12.4
12.5
3.3
2.7
3.4
0.4
9.2
0.3
0.8
1.5
3.1
3.8
1.8
0.4
0.2
0.4
0.6
0.1
1.8
2.3
2.5
2.4
0.6
2.7
0.9
1.7
1.9
2
1.3
1.9
0.9
3.4
2.3
2.5
2.1
3.7
5.1
0.3
1.1
4.9 22
25.3 23.5
37.1 22.8
10
11.3
1984
0.5
0.2
1990
0.1
5.8 21.0 42.5
66.6 67.3 50.4
9.1 29
0.4
5.8
1.3
5.6
0.1
0.9
A STATISTICAL ANALYSIS O F FOREIGN EXCHANGE RATE BEHAVIOUR IN NIGERIA'S AUTCTION
15
naira and the process by which the value is arrived at have undergone tremendous changes.
The value depreciated from a nominal rate of n1.004 to US$1,000 in mid-September
1986 to about N9.8662 to $1.000 as at the end of December 1991, and to about N18.55
to a dollar through deliberate depricaiton on 5 March 1992 by the federal government.
The auction system has equally undergone several changes during this time. Between
September 1986 and November 1991, the auction system passed through six major phases
(Table 3). A much finer classification may put these phases into seven by splitting the
frequency of bidding for the interval from March 1991 to November 1991 into two parts
namely, twice weekly in the first instance, and then once weekly.
Table 3: Nigeria's foreign exchange market phases
Period
Phase
Phase
Phase
Phase
one
two
three
four
Phase five
Phase six
26 Sep 1986 - 2 Oct 1986
2 Oct 1986 - 19 Mar 1987
19 Mar 1987 - 1 2 Dec 1988
9 Jan 1989 - 14 Dec 1990
14 Dec 1990 - 1 9 Mar 1991
Mar 1991 - Mar 1992
Auction system
Frequency of biddings
Average rate pricing
Marginal rate pricing
Dutch Auction
Inter-bank market
(with ruling rates
determined by the CBN
using some guiding
principles)
Dutch auction
Dutch auction
Weekly
Weekly
Fortnightly
Daily
Daily
Weekly
Source: From data obtained from the Central Bank of Nigeria.
Phases one and two: September 1986 - March 1987
The first phase of the auction system lasted only from 26 September 1986 to 2 October
1986 - two auctions. Under this phase, bidding was done weekly and the exchange rate
was arrived at through the averaging of all bids submitted. This phase has come to be
known as a phase of average rate pricing.
The method of determining the exchange rate under the average rate pricing system
was modified after the second auction on 2 October 1986. This modification resulted in
the second phase of the auction system, which was based on marginal rate pricing principle.
The frequency of bidding was retained however. This phase of the auction system lasted
until 19 March 1987.
The first and second phases of the auction system, fell within the period in which
Nigeria began a major reform programme. The macroeconomic policies of the
government continued the National Economic Recovery Programme that had taken off
in October 1985. There was upward adjustment in customs and excise tariffs and a ban
was placed on the importation of specific commodities, including vegetable oils,
commercial day-old chicks, etc. Non-statutory transfers to all economic and quasi-
16
RESEARCH PAPER 4 9
Table 4: Federal government budget deficits (1986-1990)
Year
Deficit
(Nbn)
GDP at
Total
Deficit
Current
expenditure
as % of
% of total
GDP
expenditure
prices ( N b n )
(Nbn)
Deficit as
1986
1987
1988
1989
8.25
5.89
12.16
15.27
73.06
108.88
145.23
196.16
16.22
22.02
27.75
41.03
11.29
5.41
8.37
7.78
50.86
26.75
43.82
37.22
1990
23.36
239.79
64.15
9.74
382.00
economic parastatals were reduced by about 50% and subsidy on petroleum products,
excepting household kerosene, was cut to about 20%. The fiscal measures led to a 25.5%
reduction in the fiscal deficit by the end of 1986.
Monetary and credit policies adopted in 1986 were aimed at reducing excess demand
and liquidity in the economy. Overall net domestic credit was expected to increase by
\ 5% and 6% in 1986 and 1987, repectively.
In actual performance, however, the governtment budget experienced a deficit of
about 11.3 % in 1986 — which was substantially above the 3 % envisaged for the structural
adjustment programme (SAP) and 5.4% in 1987 (Table 4). Bank credit to the economy
rose by 12.7% between 1985 and 1986 instead of the 8.7% allowed by the programme.
The corrresponding figures for 1987 and 1988 were 14.3% and 36.2% increases as
compared with 6.0% and 8.1% projected for the SAP.
The depreciation experienced in the dollar exchange rate during the first phase could
have been caused partly by the inability of the government to mainatain fiscal and
monetary targets, while the appreciation in the second phase could be assumed to have
resulted from the massive reduction in fiscal deficits from 11.3% of GDP to 5.4% in
i 1987, even though there was considerable overshooting of the credit target.
Phase three : March 1987 - December 1988
The period 19 March 1987 and 12 December 1988 marked the third phase in the exchange
rate management. During this phase, the Dutch auction was adopted and the frequency
of bidding was changed from once weekly to once fortnightly.
The massive depreciation experienced by the dollar exchange rate over this phase
can be traced to the liberal policies the government pursued, especially in 1988. In 1988,
the government implemented a reflationary programme to increase the level of economic
activity, which had reached a very low ebb during the second phase of the auction system.
The reflationary budget raised the deficit from 5.4% in 1987 to about 8.4% in 1988. The
specific fiscal measures designed to reflate the economy consisted of the following:
A STATISTICAL ANALYSIS O F
FOREIGN EXCHANGE RATE BEHAVIOUR IN NIGERIA'S AUTCTION
17
Provision of a reflationary package of N2.5 billion in addition to the built-in
deficit of N6 billion, thus raising the overall budget deficit to N8.5 billion
•
Exemption from taxation of all investment income earned outside Nigeria and
repariated to Nigeria through official channels.
•
Adopttion of lower company tax rates for three years for small and mediumsize companies (annual turnover of less than N500,000) engaged in
manufacturing,agricultural production or mining.
•
Revocation of the wage freeze and restoration of the right to collective
bargaining, including the adoption of enchanced wages/salaries and fringe
benefits package in the public sector.
Prescription of an additional 5 % capital tax allowance for qualified expenditure
on plant construction and agricultural production.
These fiscal measures coupled with the tremendous increase in credit from a target of
8.1% to 36.2% could have exerted substantial pressure on the exchange rate, which
reached the highest level of depreciation over all the six phases.
9
Phase four: January 1989 - December 1990
The fourth phase in the auction system spanned about two years, 9 January 1989 to 14
December 1990. This phase introduced the inter-bank foreign exchange market system
with daily biddings for foreign exchange by commercial banks.
During this phase, the dollar exchange rate experienced about 18.9% depreciation.
This was clearly an "improvement" on the 24.9% experienced during the third phase.
There are at least two major policy factors that contributed to the reduction of the pressure
on the naira. The first is the marginal reduction in the budget deficit from 8.4% in 1988
to 7.8% in 1989. But the impact of this benefit was later wiped out by the 9.7% deficit
incurred in 1990 budget. The second major factor was the directive in 1989 that all
federal and state government establishments should move all their accounts from
commercial banks to the Central Bank.
Other monetary policy measures adopted in 1989 included:
Continued deregulation of interest rates, which raised the CBN rediscount rate
to 13.25%.
Increase in the cash reserve ratio by 1%.
Increase in the statutory minimum liquidity ratio from 27.5% to 30%.
Growth in aggregated credit, targeted at 9.5% actually declined by 14.1% by the end
of the year. However, money supply which was expected to grow at 14%, shot up by
21.5%.
The rather high level of depreciation of the dollar exchange rate during this phase
may have been caused by the increased budget deficit experienced in 1990 and the
enormous increase in money supply. The increase in money supply projected at 13.0%
rose as far as 44.9% by the end of the year, while the aggregate credit target, set at 13.5%
stood at 17.1% at the end of 1990.
9
9
9
18
RESEARCH PAPER 4 9
Phase five December 1990 - March 1991
Between 14 December 1990 and mid-March 1991, the Dutch auction was once again
introduced, but with daily biddings. This marked the fifth phase of auction system.
This phase fell into the first quarter of 1991 and the macroeconomic policies of the
government were just being implemented. A modest budget surplus of MO. 1 million was
expected and the growth in aggregated bank credit was pegged at 10.6% with all creidt
going to the private sector. Money supply was also expected to grow at 14.6%. The very
low level of depreciation of the dollar exchange rate (3.62%) could have resulted from
the fact that the economy was yet to feel the impact of public sector expenditure
prgorammes and policies whose implementation had just commenced. Probably such
impacts start appearing on the economy when spending agencies of the government
receive the first allocation of their respective budgets and when the implementation of
new policy directives has actually taken off. For instance, the impact of the slash of
lending rates to a maximum of 21% in January 1991 is not likely to show up in the
demand for credit until later in the year.
Phase six: March - December 1991
From mid-March 1991 to early March 1992, the naira exchange rate was determined by
means of the Dutch auction, but the frequency of bidding was changed from daily to
weekly. The auction system was thus in its sixth phase before its collapse on 5 March
1992.
The increase deprectiation and variability experienced in the dollar exchange rate
over this phase could have been the result of pressure on the naira arising from the huge
budget deficit that the government ran instead of the expected surplus (Table 5) and the
inability to mainatain set targets in the area of monetary and credit policy.
The movements in the principal exchange rate since commencement of the marketdetermined exchange rate management system in the country could, therefore, have been
influenced substantially by the type of macroeconomic policies the government pursued
over the various phases of the auction system. In the next section, we take up the analysis
of the effects that, together with the policy characteristics, could exert more extensive
impact on the variability of the naira exchange rate.
Behaviour of the naira - dollar exchange rate under the
auction system
Performance of the exchange rate under each phase
The shortest phase in the auction system was the first one, while the longest was the
19
A STATISTICAL ANALYSIS O F FOREIGN EXCHANGE RATE BEHAVIOUR IN NIGERIA'S AUTCTION
Table 5:
1991 budget performance (Nbn)
1991 budget
out-run
1991 approved
budget
Items
Retained revenue
Recurrent
expenditure
Capital (treasury
component)
Debt service
Total expenditure
Surplus deficit
Difference
38.766
27.575
-11.191
12.334
12.484
0.15
9.69
16.642
38.666
0.100
9.943
24.588
47.012
-19.437
0.253
7.943
8.246
-19.537
Source: National Planning Commission.
Table 6: Summary of naira-dollar exchange rate index
Auction
system
Mean Variance Rate of
Mean
Variable
Frequency
of
rate of of rate of dep/app
of biddings exchange
rate
exchange change change
rate
N
%
GOV Instability
index
%
%
%
3.9235
8.7182
0.0100
Average rate
pricing
Weekly
4.8379
0.0486
1.9808
Marginal rate
pricing
Weekly
3.7159
0.1773
-0.2615
27.1493 ••25.8200
0.0477
Dutch auction
Fortnightly 4.2970
0.1320
0.3148
4.2216 24.8963
0.0307
market
Daily
7.7308
0.1116
0.0185
0.0234
18.8873
Dutch auction
Daily
9.1696
0.1604
0.0299
1.1168
3.6231
Dutch auction
Weekly
10.0668
0.7467
0.1139
0.0975
10.8813
0.3103
Inter-bank
Source: Computed from data obtained from Central Bank of Nigeria
Note: 1. Figure in absolute terms.
2. Coefficient of variation.
fourth. The mean value of the exchange rate was highest under the last phase (Dutch
auction with weekly biddings) and was lowest under the second (marginal rate pricing
with weekly biddings), during which the exchange rate appreciated substantially (25.8%)
(Table 6). From the mean value of the rates of change, the exchange rate moved least
under the inter-bank market (phase four). It had a mean of change of 0.0185% and a
20
RESEARCH PAPER 4 9
variability index of 0.0234. The index of instability, which synthesizes the effects of the
variance in the exchange rate and the rate of change of the exchange rate, reported an
index of 0.0454 for the inter-bank market. This ranks this phase as most stable. This
phase was followed closely by the results of the indexes recoreded for phase five (Dutch
auction with daily biddings). Under this phase the exchange rate had a mean rate of
change of 0.0299% and a variability index of 1.1168. Its instability index is 0.0693.
Ruling out phase one, since this marks the period during which the first shock was
introduced into exchange rate management policy in the country, it is obvious from Table
6 that exchange rate movement was most volatile under phase six (Dutch auction, weekly
bids), followed by phase two (marginal rate pricing) and phase three (Dutch auction with
fortnightly bids). With respect to the rate of depreciation or appreciation, the exchange
rate was relatively more stable under the Dutch auction with daily bids (phase five). It
depreciated most substantially under phase three — Dutch auction with fortnightly bids.
It was only phase two, marginal rate pricing, that generated appreciation in the exchange
rate.
From the point of view of the naira-dollar exchange rate behaviour, the results indicate
that a stable and realistic exchange rate for the Nigerian naira is more likely to be generated
through the inter-bank market and Dutch auction with daily biddings.
Relative dispersion of exchange rate bids across buyers
in the foreign exchange market
In the first phase of the auction system which spanned the period 26 September - 10
October 1986, there was a wide gap between the highest and lowest bid rates submitted
by dealers operating in the market. The highest bid rate in the first auction was N5.1250
to US$ 1, while the lowest rate quoted was N2.5000. Out of a total of 34 banks participating
in the auction, only 16 were successful. Since no bank was disqualified, it means 18
banks quoted rates that were far below the average rate of N4.6406 at which the Central
Bank sold foreign exchange to the authorized dealers. Thus, about 50% of buyers during
the first auction submitted bids that fell on both sides of the average. The distribution of
bid rates under the first auction was normal.
During the second auction, which marked the end of the first phase, the gap between
the highest and lowest bid rates was maintained. The rates stood atN5.5999 andN3.0000.
The average rate at which authorized dealers bought foreign exchange from the Central
Bank was N5.0839. Only 13 of the 38 dealers were successful. Thus, 25, representing
over 70% of bidders, submitted bids that were much lower than the average rate at which
the CBN sold foreign exchange to the dealers. There was, therefore, a very heavy
concentration xrf bids below the average, indicating negative skewedness in the
distributiion of bid rates quoted by buyers.
The change to the marginal rate pricing system (i.e., phase two), may have been
necessitated by the need to increase the access of a much larger percentage of buyers to
foreign exchange. During this phase, the rate at which the Central Bank sold foreign
A STATISTICAL ANALYSIS O F FOREIGN EXCHANGE RATE BEHAVIOUR IN NIGERIA'S AUTCTION
21
exchange to dealers fell considerably. The highest bid rate also fell significantly, thus
reducing the gap between the highest and lowest. The near convergence of rates under
this phase increased the number of successful banks in the auctions. On the average the
highest bid rate was 4.2395 while the lowest was 3.4343, and about 25% of buyers
submitted bids that fell below the marginal rates set by the CBN. Therefore, the
distribution of the bids was postively skewed.
From the second to the sixth phases of the auction system, the gap between the highest
and lowest bid rates narrowed considerably and the number of successful banks in the
auction came close to 100%. There was, therefore, increased concentration of biddings
around the rates at which the Central Bank sold foreign exchange to authorized dealers
(Table 7).
Table 7:
Auction
system
Dispersal of biddings in the foreign exchange market
Average highest
bid rate
Average rate
pricing
Marginal rate
pricing
Dutch auction
Inter-bank market
Dutch auction
Dutch auction
Average lowest
bid rate
Average Average number
FEM rate of participating
banks
Average number
of participating
banks
5.36245
2.75
4.86225
36
14
4.23956
4.4354
7.874398
9.2404
3.43433
4.23153
7.494794
8.970145
3.65217
4.31839
7.786987
9.14854
39
48
86
107
34
43
86
104
9.864037
9.96878
111
107
10.49323
Behaviour of exchange rates for other currencies
The analysis above relates to the behaviour of the rate of change of the lead currency,
i.e., the US dollar. For the other five currencies, consistent data were available for only
four of the six phases of the auction system. These phases are the Dutch auction with
fortnightly biddings (phase three), inter-bank with daily biddings (phase four), Dutch
auction with daily biddings (phase five) and Dutch auction with weekly biddings (phase
six). The dollar is also included in this analysis for ease of comparison. Its indexes may
not correspond with those in Table 6, however, because of the data gap experienced for
some of the currencies. The summary indexes for each of the four phases for all the
currencies are reported in Tables 8 - 1 1 . The summary indexes compare mean values
and variability of the rates of change and the rates of depreciation or appreciation of
exchange rates for all currencies during each of the four phases of the auction system.
22
RESEARCH PAPER 4 9
Summary indexes for Dutch auction (fortnightly
biddings) 1988
Over this phase of the auction system, the naira exchange rate for the British pound
sterling (S£) fluctuated most; it had a mean rate of change of 0.9312%, a vairiance of
7.1016 and a rate of depreciation of 22.6872% (Table 8). Its coefficient of variation
stood at 0.078 while the exchange rate index of instability was 0.775. This was followed
by the exchange rate for the dollar, which had an instability index of 0.2619. The exchange
rate for the yen was most stable. This is clearly indicated by the coefficient of variation
and instability index.
Table 8: Summary indexes for Dutch auction (fortnightly biddings) 23 June 1988 - 8 December
1988
Currency
Variation of
exchange
rate
Mean
exchange
rate
(N)
US dollar
British pound
German mark
Swiss franc
French franc
Dutch gilder
Japenese yen
0.087747
0.645165
0.049636
0.067301
0.003946
0.039176
0.000048
4.785383
8.2962
2.638383
3.144308
0.775941
2.33875
0.035191
Rate of
Mean Variation
of
change
depreciation/
change appreciation
(%)
0.7816
0.9312
0.8225
0.0768
0.7951
0.8232
0.9287
(%)
1.7962
7.1016
2.9331
3.1489
2.7089
2.92
2.9141
COV
(%)
19.4246
22.6872
20.3297
19.1257
19.7237
20.3442
22.6328
Instability
index
(%)
0.018336
0.077766
0.018813
0.021404
0.005085
0.016751
0.001364
0.261884
0.775098
0.202053
0.071894
0.056013
0.179582
0.006677
Source: Computed from data obtained from Centrald Bank of Nigeria.
Summary indexes for inter-bank market (daily biddings)
9 January 1989 - 14 December 1990
Under the inter-bank market, the dollar exchange rate performed most remarkably well
(Table 9). With a mean rate of change of 0.0234%, a variability index of 0.0305 and a
rate of deprecitaion of 1.8873%, the naira exchange rate could be taken to have attained
a reasonably stable level. The coeffiecient of variation and the index of instability,
however, place the performance of the dollar next to the yen, which most studies have
confirmed to be most stable. But there is more to it than meets the eye. If the variability
index is an indication of the extent to which the rate of change in the exchange rate is
generated by a stochastic process (which the auction system is expected to be), then a
variability index of 0.0305 suggests an auction system that was controlled by the monetary
authorities.
A STATISTICAL ANALYSIS O F
FOREIGN EXCHANGE RATE BEHAVIOUR IN NIGERIA'S AUTCTION
23
Table 9: Summary index for inter-bank foreign exchange market 9 January 1989 - 1 4
December 1990
Currency
Variation of
Mean
exchange exchange
rate
rate
m
US dollar
British pound
German mark
Swiss franc
French franc
Dutch gilder
Japenese yen
1.130323
2.225706
0,367032
0.556702
0.081178
0.309482
0.000016
Mean
change
{%)
7.707596
0.0234
13.15465
0.0337
0.0467
4.448585
5.15046
0.0471
1.334631 -0.1292
3.9331169 0.0686
0.019
0.054175
COV
Rate of
Variation
depreciation/
of
change appreciation
Instability
index
1.8873
26.0022
34.1312
43.1312
43.3903
45.842
15.6279
0.055223
0.273873
0.130921
0.161928
0.102411*
0.145707
0.000551
(%)
0.0305
0.1369
10.079
10.079
13.808
2.4752
0.2113
0.016908
0.169195
0.082505
0.108088
0.060824
0.078685
0,000295
(%)
Source: Computed from data obtained from Central Bank of Nigeria.
Notes: 1. Coefficient of variation.
2. Measured in absolute terms.
Table 10: Summary index for Dutch auction with daily biddings 14 December 1990 - March 1991
Currency
Mean
Variation of
exchange exchange
rate
rate
(N)
US dollar
British pound
German mark
Swiss franc
French franc
Dutch gilder
0.139281 9.144045
0.67324 17.68496
0.063799 6.10956
0.087987 7.163935
0.005537 1.79755
0.052198 5.42022
Japenese yen 0.00001
Mean
change
Variation
of change
(%)
(%)
0.334225
0.276082
0.247669
0.211593
0.244098
0.251573
0.068645 0.274757
COV
Rate of
depreciation/
appreciation
0.50889
0.269961
0.428571
0.398157
0.38736
0.430227
14.26561
11.93901
10.77918
9.284537
10.63234
10.93944
0.367901 11.88525
Instability
index
(%)
0.015232
0.038069
0.010442
0.012282
0.00308
0.00963
0.215757
0.431126
0.125702
0.136446
0.036764
0.114593
0.000146
0.001658
Source: Computed from data obtained from Central Bank of Nigeria.
Note: 1. Coefficient of variation.
2. Measured in absolute terms.
The pound sterling (S£) exhibited the greatest variability over the period. It had an
instability index of 0.2739 and a coefficient of variation of 0.1692. The DM and DG
moved in the same direction with about the same magnitude of change, variance and
depreciation. In relative terms, however, the DG experienced the largest rate of change
and depreciation in its exchange rate —0.0686% and 45.842%, repectively.
24
RESEARCH PAPER 4 9
Summary indexes for Dutch auction (daily biddings) 14
December 1990 -March 1991
During this brief phase of about three months of daily biddings under the Dutch auction
system, the variability of the dollar exchange rate was relatively higher than that
experienced by all other currencies except the pound sterling. The pound sterling had an
instability index of 0.4311 and a coefficient of variation of 0.0381. These indexes are
more than twice those recorded by the dollar exchange rate — 0.02158 and 0.0152,
respectively. With the exception of the Japanese yen and French franc, the variability in
the exchange rate for the other currencies as measured by the instability index ranged
between 0.1146 for the DG and 0.1364 for the SwF. The yen maintained its relative
stability in comparison with the other currencies, and was followed by the French franc.
Apart from the dollar exchange rate, which experienced a much higher level of
depreciation, the depreciation rate for all other currencies was about 10%.
Summary indexes for Dutch auction (weekly biddings)
March 1991 -October 1991
Table 11 presents summary indexes for the last phase before the collapse of the auction
system. This phase started in mid-March 1991. The Japanese yen and the US dollar had
the lowest rate of change and variability index, as well as a low rate of depreciation. The
other five currencies experienced negative rates of change that translated into fairly high
rates of appreciation. With the exception of the I T , which had 0.0438, the variability of
the naira exchange rates for the other currencies as measured by the instability index
ranged between 0.1326 for DG and 0.4223 for the S£.
Given that the dollar is the pinicipal currency in the auction, its behaviour seems to
represent more closely the nature of the demand and supply situations in the foreign
exchange market. The appreciation experienced by the other currencies is more of a
reflection of movements in the exchange rates in international foreign exchange markets.
Performance of exchange rates across the various
phases of the auction system
Tables 12-16 present comparative indexes for all the currencies across four phases of
the auction system. The idea is to enable one to reach a preliminary conclusion as to
which phase of the auction system would be more likely to generate relatively reasonable
indexes, particularly for the dollar exchange rate, in the long run. Within the context of
our analysis the desirable auction system should be the one that generates the lowest rate
of change and variability of exchange rate changes as well as the lowest rate of
A STATISTICAL ANALYSIS O F
FOREIGN EXCHANGE RATE BEHAVIOUR IN NIGERIA'S AUTCTION
25
Table 11: Summary indexes for Dutch auction with weekly biddings March 1991 - October 1991
Currency
Variation of
exchange
rate
US dollar
British pound
German mark
Swiss franc
French franc
Dutch gilder
Japenese yen
Mean
exchange
rate
(N)
0.776406 10.07229
1.742647 17.25808
0.221434 5.86167
0.22588
6.807796
0.018302 1.727022
0.172866 5.203206
0.000043 0.00734
Variation
Rate of
Mean
change of change depreciation/
appreciation
(%)
0.011384
-0.10235
-0.10186
-0.10862
-0.10486
-0.10169
0.034033
COV
(%)
(%)
1.584369
2.550665
2.699515
2.278305
2.639913
2.757701
1.87239
0.809269
-7.57919
-7.7542
-8.06145
-7.77208
-7.52872
2.4
Instability
index
0.077083
0.100976
0.0037777
0.03318
0.010597
0.033223
0.000586
0.094012
0.422306
0.150184
0.156629
0.043805
0.132584
0.00121
Source: Computed from data obtained from Central Bank of Nigeria.
Notes: 1. Coefficient of variation.
2. Measured in absolute terms.
depreciation. Also, it should have the lowest coeffiecient of variation and exchange rate
index of instability. From all indications, it seems that the path to a stable and sustainable
exchange rate based on the dollar exchange rate, is more likely to be traced by the interbank market and Dutch auction with daily biddings.
With respect to the other currencies, the relative stability of the exchange rates for the
yen and French franc is an indication that the Japanese and French monetary authorities
may have fixed their respective exchange rates.
From the foregoing analysis of the behaviour of the naira exchange rate for all the
currrencies, it is certain that a wide variation exists in cross-currency performance with
respect to the variance of exchange rate, the size of the mean rate of change, the sizes of
the coefficient of variation and instability index, and the rate of depreciation or
appreciation. This rather enormous variability in cross-currency performance tends to
show that exchange rate movements for some of the currencies are influenced by factors
outside the Nigerian economy. The case of the Japanese yen and French franc appears
well established.
Even though cross-currency variability in performance indexes is an issue of concern
in this study, the marked difference in the behaviour of each currency from one auction
system to the other is of greater consequence in the management of the market-determined
exchange rates in Nigeria. While the coefficient of variation in the exchange rates for
the dollar shows that variability is about the same across the phases of the auction system,
the exchange rate instability index, which combines variance with rate of change, shows
that the various phases of the auction system did not generate uniform levels of instability.
The absence of cross-phasal uniformity is equally shown by the mean rate of change, the
variance of the rate of change and, very importantly, the rate of depreciation/appreciation.
These differ markedly from one phase to the other in the auction system.
The cross-phasal variability in the rate of change and depreciation (appreciaton) of
the dollar exchange rate may be due to the inability of the government to maintain fiscal
26
RESEARCH PAPER 4 9
Table 1 2 : Summary of naira exchange rate indexes: all currencies mean rate of change
Currency
US dollar
British pound
German mark
Swiss franc
French franc
Dutch gilder
Japanese yen
Dutch auction
(fortnightly)
Inter-bank
(daily)
Dutch auction
(daily)
0.781647
0.931239
0.82253
0.768247
0.795104
0.823188
0.928692
0.02343
0.033706
0.046732
0.047068
-0.12915
0.068643
0.01902
0.334224
0.276082
0.247668
0.211593
0.244097
0.251572
0.274756
Dutch auction
(weekly)
0.011383
-0.10234
-0.10186
-0.10861
-0.10485
-0.10169
0.034032
Table 13: Summary of naira exchange rate indexes: All currencies variance of rate of change
Currency
US dollar
British pound
German mark
Swiss franc
French franc
Dutch gilder
Japanese yen
Dutch auction
(fortnightly)
1.796156
7.101576
2.933069
3.148895
2.708947
2.919999
2.914055
Inter-bank
(daily)
0.030478
0.136856
10.07898
10.07898
13.80804
2.475151
0.211274
Dutch auction
(daily)
Dutch auction
(weekly)
0.050889
0.269964
0.42857
0.398156
0.8736
0.430227
0.3679
1.584368
2.550665
2.699514
2.278305
2.639913
2.7577
1.872389
Table 14: Summary of naira exchange rate indexes: All currencies rate of depreciation/appreciation
Currency
US dollar
British pound
German mark
Swiss franc
French franc
Dutch gilder
Japanese yen
Dutch auction
(fortnightly)
Inter-bank
(daily)
19.42462
22.68721
20.32972
19.12574
19.72368
20.3442
22.63279
1.88732
26.00216
34.12121
34.13121
34.39034
45.84202
15.6279
Dutch auction
(daily)
14.26561
11.939
10.77917
9.284536
10.63233
10.93944
11.88524
Dutch auction
(Weekly)
0.809269
-7.57919
-7.57919
-8.06145
-7.77207
-7.52872
2.4
A STATISTICAL ANALYSIS O F
FOREIGN EXCHANGE RATE BEHAVIOUR IN NIGERIA'S AUTCTION
27
Table 15: Summary of naira exchange rate indexes: All currencies exchange rate instability
index
Currency
US dollar
British pound
German mark
Swiss franc
French franc
Dutch gilder
Japanese yen
Dutch auction
(fortnightly)
0.2619
0.7751
0.2021
0.0719
0.056
0.1796
0.0067
Inter-bank
(daily)
0.0552
0.2739
0.1309
0.1619
-0.1024
0.1457
0.0006
Dutch auction
(daily)
0.2158
0.4311
0.1257
0.1364
0.0368
0.1146
0.0017
Dutch auction
(weekly)
0.094
0.4223
0.1502
0.1566
0.0438
0.1326
0.0012
Table 16: Summary of naira exchange rate indexes: All currencies coefficient of variation
Currency
US dollar
British pound
German mark
Swiss franc
French franc
Dutch gilder
Japanese yen
Dutch auction
(fortnightly)
0.0183
0.0778
0.0188
0.0214
0.0051
0.0168
0.0014
Inter-bank
(daily)
Dutch auction
(daily)
0.0169
0.1692
0.0825
0.1081
0.0608
0.0787
0.0003
0.0152
0.0381
0.0104
0.0123
0.0031
0.0096
0.0001
Dutch auction
(weekly)
0.0771
0.101
0.0378
0.0332
0.0106
0.0332
0.0006
discipline and keep to monetary targets. It could equally be the result of day-of-theweek or month-of-the-year, auction system, and frequency of bidding effects. The
potential impacts of such effects are examined in the next section.
VI. Analysis of effects
Given the concern for the non-uniformity in the rate of change and depreciation of the
naira-dollar exchange rate, this tudy examined the potential impact of the following
effects as they relate to the relative frequency of the deprication of the exchange rate:
Auction system effect
Frequency of bidding effect
Auction system cum frequency of bidding effect
Month-of-the-year effect
Day-of-the-week-effect
With respect to the auction system, the six approaches that were tried out before the
collapse of the bidding system in March 1992 were grouped into four phases, namely,
average rate pricing, marginal rate pricing, Dutch auction and inter-bank market phases.
This grouping aimed to eliminate duplication of the Dutch auction system which could
be differentiated by means of frequency of bidding (see Table 3). For frequency of
bidding, the analysis was based on all three varieties used — weekly, fortnightly and
daily biddings. For the combined effect of auction system and frequency of bidding, all
six phases of the auction system in Table 3 formed the basis of analysis.
In examining month-of-the-year effect, data for the period 1986-1991 (Table 17) were
used, while for the day-of-the-week effect, daily exchange rate data for January - December
1990 were used. The data traversed sections of the two main phases in the auction
system that featured daily biddings, the inter-bank and Dutch auction. All analyses were
done using the chi-square and, where applicable, the moment coefficients of skewedness
and kurtosis.
8
8
8
8
8
Auction system effect
In Table 18, we report the results for the first three effects. The null hypothesis against
which the chi-square test was performed was that the depreciation occurring in each of
the phases of the four auction systems was independant of these auction systems. The
result in the table shows that at the 1% level of significance, this hypothesis is rejected.
The level of depreciation the naira-dollar exchange rate experienced before March 1992
was significantly influenced by the auction systems the monetary authorities adopted.
A STATISTICAL ANALYSIS OF FOREIGN EXCHANGE RATE BEHAVIOUR IN NIGERIA'S AUTCTION
Table 17:
Month
January
February
March
April
May
June
July
August
September
October
November
December
29
Monthly naira-dollar exchange rate movements; January - December
1986
4.62
4.12
3.54
3.19
1987
1988
1989
1990
1991
2.71
3.08
3.48
3.48
3.51
3.73
3.81
4
4.21
4.28
4.3
4.17
4.18
4.27
4.32
4.21
4.11
4.2
4.61
4.49
4.72
4.78
5.15
5.36
7.75
7.39
7.59
7.58
7.51
7.35
7.2
7.26
7.35
7.4
7.51
7.63
7.87
7.9
7.94
7.94
7.94
7.95
7.94
7.96
7.98
8.01
8.32
8.71
9.33
9.52
9.98
0.95
9.25
10.03
10.6
11.62
9.96
9.86
9.86
9.87
Number of annual biddings in the foreign exchange market.
Frequency of bidding effect
Results of the chi-square analysis for frequency of bidding are also reported in Table 18.
The evidence shows that fortnightly, weekly and daily frequencies of bidding do not
generate uniform levels of depreciation. The null hypothesis of uniform depreciation is
rejected even at the 1 % level. The depreciation experienced in the naira-dollar exchange
rate is not independent of the frequency of bidding in the auctions. The impact of the
frequency of bidding is less than half of the effect associated with the auction system.
Auction system cum frequency of bidding effect
The combined effect of auction systems cum frequency of bidding generate a much
higher value for the chi-square (see Table 18). This shows that the rate of depreciation is
accelerated when the naira-dollar exchange rate is determined through an auction system
and frequency of bidding, which are prone to depreciation. Expectedly, the effect of the
auction systems cum frequency of bidding is greater than that of each taken singly. There
is, therefore, evidence of mismatch of auction system and frequency of bidding, which
contributed to the rapid depreciation of the naira. The auction system together with the
corresponding bidding frequencies generated rates of depreciation that are significantly
different at the 1% level.
30
RESEARCH PAPER 4 9
Table 18: Auction system and frequency of bidding effect on depreciation/appreciation
Auction system bidding
frequency
Chi-square
coefficient
Auction system
Frequency of bidding
Auction system and
frequency of bidding
Critical values
D.F.
(1%)
(5%)
121.038
11.34
7.815
6.251
3
54.354
9.21
5.991
4.605
2
128.673
15.09
9.236
5
11.07
(10%)
Month-of-the-year effect and nature of distribution of
exchange rate changes
As part of the analysis of effects, this study examined the average of depreciation and the
nature of the distribution of the exchange rate changes for each month of the year for the
period September 1986 - December 1991. The results of the analysis are reported in
Table 19.
A close look at the average rate of depreciation shows that there is a fairly uniform
rate of depreciation and exchange rate changes across the months of the year. The highest
depreciation seems occurred in December (21.46%), while the lowest rates were
experienced in October (17.845%). The gap between the two is less than 4 percentage
points. In the content of our analysis, this is negligible.
This observation is supported by the result of the chi-square test for the relationship
between month of the year and rate of depreciation. Acoefficient of 0.6132 was obtained,
which is far lower than the critical value of 5.578 at the 10% level. We cannnot, therefore,
reject the hypothesis that, generally, the depreciation the naira-dollar exchange rate
experiences on a monthly basis is independant of the month of the year. In other words,
the month of the year does not seem to exert significant effect on either the rate of
depreciation or the change in the exchange rate.
Given the empirical evidence, however, the data in the table also point to a fact that
has been observed with respect to exchange rate behaviour in Nigeria. That is the exchange
rate tends to depreciate more in December and January than all other months, which the
results in Table 19 confirm a relatively higher level of depreciation also tends to occur
about the middle of the year. The results point to the month of July.
The distribution of the rate of change shows positive skewedness in all cases. The
peaks of the distribution as measured by the coefficient of kurtosis show that the months
of January, and July-December exhibited platykurtic distribution, which suggests greater
dispersion of the rate of change, while the months of February-June showed leptokurtosis
in their distribution. The February-June distribution, therefore, indicates greater
peakedness or less variability in the exchange rate changes. The exchange rate changes
are therefore not normally distributed.
A STATISTICAL ANALYSIS O F FOREIGN EXCHANGE RATE BEHAVIOUR IN NIGERIA'S AUTCTION
31
Table 19:
Month-of-the-year effect and the nature of distribution of the rate of exchange rate
changes 1986 - 1990
Month
January
February
March
April
May
June
July
August
September
October
November
December
Average rate Average rate of
of depreciation exchange rate
changes
Nature of distribution
Skewedness
Kurtosis
-0.03186
21.28589
7.638211
1.510002
20.65866
7.17532
1.721716
0.004226
18.89488
6.518969
2.021696
0.145588
18.71854
6.565209
2.121878
0.194962
16.48668
6.100275
2.087568
0.17713
17.93409
6.524655
2.030566
0.154634
21.01425
7.333039
0.781204
-0.43415
18.884
6.608528
1.569723
-0.01202
18.16454
4.702369
0.800164
17.84504
5.772881
1.662168
-0.02013
19.31359
7.449067
0.120842
-0.74227
21.46121
8.741283
0.222483
-1.03816
-0.40475
Pos tive
Pos tive
Pos tive
Pos tive
Pos tive
Pos tive
Pos tive
Pos tive
Pos tive
Pos tive
Pos tive
Pos tive
platykurtic
leptokurtic
leptokurtic
leptokurtic
leptokurtic
leptokurtic
platykurtic
playtkurtic
platykurtic
platykurtic
platykurtic
platykurtic
Chi-square coefficient for month-of-the year effect on depreciation: 0.613207
Critical values:
3.053 (1%), 4.575 (5%), 5.578 (10%).
Day-of-the-week effect and nature of distribution of
exchange rate changes
The final effect that was examined in this study concerns the impact of the day of the
week on depreciation, and the nature of the distribution of exchange rates changes for
each of the week. Like the analysis above day-of-the-week tests were carried out to see
whether nor not the IID properties of the daily exchange rates can be confirmed or rejected.
Results of the test are reported in Table 20.
Looking closely at the average rate of depreciation and rate of exchange rate changes
for each day of the week, we observed that the naira-dollar exchange rate tends to
depreciate more on Mondays and Thursdays. It appreciates on Tuesday, Wednesday and
Fridays. This trend is also noticeable in the average rate of exchange rate changes for
each of the week. This rules out indentical rates of depreciation across days of the week.
The highest depreciation seems to occur on Thursdays (0.3716%), while the highest
appreciation appears to take place on Wednesdays (0.0588%).
The rejection of identical rates depreciation is confirmed by the result of the chisquare test for the relationship between the day-of-the-week and the rate of depreciation.
The empirical investigation turned out a chi-square coeffieient of 52.027, which is
considerably higher than the critical value of 13.28 (1% level of significance). We
therefore reject the hypothesis that, generally, the depreciation the naira-dollar exchange
rate experiences on a daily basis is independent of the day of the week in which foreign
32
RESEARCH PAPER 4 9
Table 20: Day-of-the-week effect and the nature of distribution of the rate of exchange rate
changes 1986-1990
Day
Monday
Tuesday
Wednesday
Thursday
Friday
Average rate Average rate of
of depreciation exchange rate
changes
0.269542
0.098607
0.058776
0.371648
0.068634
0.120042
0.042926
0.025547
0.165299
0.02986
Nature of distribution
Skewedness
Kurtosis
29.57145
10.77053
2.817027
20.5572
0.253696
28.37688
11.85152
2.769936
21.351
0.253696
Positive,
Positive,
Positive,
Positive,
Positive,
Leptokurtic
Leptokurtic
Leptokurtic
Leptokurtic
Leptokurtic
Chi-square coefficient for day of the week effect on depreciation: 52.02736
Critical values: 13.28 (1%), 9.488 (5%), 7.779 (10%).
exchange transactions are carried out. In other words, in assessing factors that lead to
the depreciation of the naira-dollar exchange rate, there is need to adjust the exchange
rate data series for the impact exerted by the day of the week.
VII. Summary and conclusion
Exchange rate management policy in Nigeria has passed through five major stages, the
current one of which is characterized by market-determined exchange rate management.
The current system, which began on 29 September 1986, constitutes a fundamental aspect
of the country's economic reform programme. Since the onset of this approach, one
major problem that has confronted the Nigerian monetary authorities has been the marked
instability in the principal exchange rate — the naira-dollar rate. The instability has
mainly featured depreciation of the rate.
In the effort to establish a stable and variable exchange rate for the naira, the monetary
authorities experimented with four auction systems and six frequencies of bidding between
September 1986 and March 1992 before the collapse of the bidding system. The behaviour
of the dollar exchange rate differs significantly over these auction systems and bidding
frequencies. This applies equally to the other six currencies used for the study. We
found, however, that the dollar exchange rate was relatively stable under the inter-bank
(IFEM) and Dutch auction systems with daily frequencies of bidding. The Japanese yen
and French franc were most stable throughout all phases of the auction system. This
gives the impression that the Japanese and French monetary authorities must have fixed
their exchange rates. The British pound sterling exhibited greater characteristics of
instability.
In the attempt to explain some of the possible causes of the instability, particularly in
the principal currency, our study looked at the macroeconomic policy characteristics of
all six phases of the auction system between September 26 1986 and December 1991.
We found the fiscal and monetary discipline was lacking on the part of the government,
but this does not sufficiently account for the steep depreciation the naira has experienced.
We therefore examined the impact of the five major factors on the rate. These are the
auction systems, effect frequency of bidding effect, auction system cum frequency of
bidding effect, month-of-the-year effect, and day-of-the-week effect.
For the auction system effect, we found that the level of depreciation the naira-dollar
exchange rate experienced was significantly influenced by the type of auction system
the monetary authorities adopted. We also found that frequency of bidding exerted some
influence on the rate of depreciation. The joint impact of both the auction system and
frequency of bidding was much larger on the rate of depreciation when compared with
their individual influences. On the month-of-the-year effect and the nature of the
distribution of exchange rate changes, we found that the highest level of depreciation
tends to occur in December and January, but generally the depreciation the dollar exchange
34
RESEARCH PAPER 4 9
rate experiences is not significantly dependant on the month of the year, and the exchange
rate changes are not normally distributed.
Finally, with respect to the day-of-the-week effect and nature of distribution of
exchange rate changes, it was found that the depreciation that occurs in the daily biddings
is not independent of the day of the week. The dollar exchange rate tends to appreciate
on Tuesdays, Wednesdays and Fridays. The highest depreciation seems to occur on
Thursdays, while the highest appreciation appears to take place on Wednesdays.
Given these findings we wish to conclude that to adequately account for the instability
of the naira exchange rates, it is desirable to look at properties of the processes that
generate the exchange rates, effects from day of the week and the macroeconomic policy
environment among other factors. The search for the relative significance of these factors
in the movement of the exchange rate over the time will be left for the second phase for
this research.
VIII. Policy implications
The policy implications that come out of this study are threefold. These are as follows:
1.
2.
3.
The monetary authorites should adopt one mechanism for determining the
exchange rate, having tried out six pricing/bidding systems. Frequent changes
in the auction system and frequencies of bidding should be avoided. Besides,
given the time series properties of the exchange rate data, it is not unlikely
that the exchange rate will not return to its equilibrium level once disturbed.
A perfectly stable exchange rate under the market-determined exchange rate
management system cannot be achieved, because of the response of monthof-the-year and the day-of-the-week effects. This means that not all
variations in the exchange can be eliminated. The monetary authorities should
estimate the size of these effects and make allowance for them when
determining the bounds over which the exchange rate may not be allowed to
vary, once a particular method for determining the rate has been established.
The federal government should reduce its fiscal deficit so that effective
monetary restraint can be exercised by the Central Bank. It is common
knowledge that a larger part of the excess fiscal burden of government
operations is borne by the monetary sector.
References
Boothe, P. and D. Glassman. 1985. "The statistical distribution of exchange rates:
empirical evidence and economic implications " ". Unpublished Manuscript (U.S. A,
Alberta: Department of Economics, University of Alberta).
Burt, J., F. R. Kaen and G. G. Booth. 1977. "Foreign market efficiency under flexible
exchange ratesJournal of Finance, vol. 32. pp. 1325 - 1330.
Calderon - Rosell, J. R. and M. Ben-Horim. 1982. "The behaviour of foreign exchange
rates" Journal of International Business Studies, vol. 13. 1982. pp. 99 - 111.
Elbadawi I. A.1991. "The black market for foreign exchange and macroeconomic
management in Sudan" . Mimeo (U.S.A.; Washington: The World Bank, May).
French K. R. 1980. "Stock returns and the weekend effect". Journal of Financial
Economics vol. 8. pp. 55 - 69.
Friedman, D. S. Vandersteel. 1982. "Short-run fluctuations in foreign exchange rates".
Journal of International Economics, vol. 13. pp. 171 - 186.
Gibbons, M. R. and P. Hess. 1981. "Days of the week effects and asset returns". Journal
of Business, vol. 54. pp. 579 -596
Hsieh, D. A. 1988. "The statistical properties of daily foreign exchange rates: 1974 1983". Journal of International Economics, vol. 24. pp. 129- 145.
Hogg, R. V. and A. T. Craig. 1970. "Introduction to Mathematical Statistics". (New
York: macmillan).
Keim, D. B. and R.F Stambaugh, 1984. "A further investigation of the weekend effect in
stock returns", Journal of Finance, vol. 39. pp. 819 - 837.
Messe, R. A. and A. K. Rose. 1991. "An empirical assessment ofnonlinearities in models
of exchange rate determination". The Review of Economic Studies, vol. 58 (3).
no. 195.
National Bureau of Economic Research Inc. 1991. "Pitfalls and opportunities: What
macroeconomists should know about Unit Roots". Unpublished Manuscript
(U.S.A., Cambridge Massachusetts).
Omoruyi, S. E. 1989. "Structural adjustment and national development; Deriving a fair
and viable exchange rate for the Naira ". Central Bank of Nigeria, Economic and
Financial Review. September, pp. 29 - 34.
Stock, J. H. and M. Watson. 1988. "Variable trends in economic time series". Journal
of Economics Perspectives, vol. 2. pp. 297 - 232.
United Bank of Africa. 1992. Monthly Business & Economic Digest. (Nigeria: Lagos,
March).
A
A STATISTICAL ANALYSIS O F FOREIGN EXCHANGE RATE BEHAVIOUR IN NIGERIA'S AUTCTION
37
Westerfield, J. M. 1977. "An examination of foreign exchange risk under fixed and floating
rate regimes". Journal of International Economics, vol. 7. pp. 181 - 200.
AFRICAN ECONOMIC RESEARCH CONSORTIUM
P.O. BOX 62882
NAIROBI, KENYA
TELEPHONE (254-2) 228057
225234 215898 212359
332438 225087
TELEX 22480
FAX (254-2) 219308
E-MAIL [email protected]
T h e principal objective of the African Economic Research
Consortium (AERC), established in August 1988, is to
strengthen local capacity for conducting independent, rigorous
inquiry into problems pertinent to the management of
economies in Sub-Saharan Africa.
In response to special needs of the region, AERC has
adopted a flexible approach to improve the technical skills of
local researchers, allow for regional determination of research
priorities, strengthen national institutions concerned with
economic policy research, and facilitate closer ties between
researchers and policy makers.
Since its establishment, AERC has been supported by
private foundations, bilateral aid agencies and international
organizations.
SPECIAL PAPERS contain the findings of commissioned
studies in furtherance of AERC's programmes for research,
training and capacity building.
RESEARCH PAPERS contain the edited and externally
reviewed results of research financed by the AERC.
It is AERC's policy that authors of Special and Research
Papers are free to use material contained therein in other
publications. Views expressed in the Special and Research
Papers are those of the authors alone and should not be
attributed to the AERC's sponsoring Members, Advisory
Committee, or Secretariat.
Further information concerning the AERC and additional
copies of Special and Research Papers can be obtained by
writing to: African Economic Research Consortium, P.O. Box
62882, Nairobi, Kenya.
ISBN 9966-900-74-8
This work is licensed under a
Creative Commons
Attribution - Noncommercial - NoDerivs 3.0 Licence.
To view a copy of the licence please see:
http://creativecommons.org/licenses/by-nc-nd/3.0/